Networking and Communication Engineering

S. Bala Manohari ¹, R. Sasikala ¹

Affiliations
1 Department of Information Technology, K.S.R. College of Technology, Tiruchengode, Tamil Nadu, IN

Abstract

Over the last couple of years, “Cloud Computing” has emerged as a compelling and successful paradigm for internet scale computing. It depends primarily on IaaS layer to provide cheap and pay-as-you-go processing power, data storage, and other shared resources. Scalable database management systems - both for update intensive application workloads as well as decision support systems for descriptive and deep analytics are a critical part of the cloud infrastructure and play an important role in ensuring the smooth transition of applications from the traditional enterprise infrastructures to next generation cloud infrastructures. Cloud computing platforms provide scalability and high availability properties for web applications but they sacrifice data consistency at the same time. The proposed work, a Transaction Processing System, guarantees full ACID properties for multi-item transactions issued by Web applications, even in the presence of server failures and network partitions by the use of Local Transaction Manager. The TPS provides both strong consistency guarantees and high availability. For each type of databases stored in the cloud separate LTMs are created for handling it. The particular LTMs will take care of the requests generated for the database linked with it and in turn it will increase the efficiency and consistency of the system.

Keywords

Scalability, Web Applications, Cloud Computing.

Full Text

     

R. Sasikala ¹, K. Duraiswamy ²

Affiliations
1 K. S. Rangasamy College of Technology, Tiruchengode–637 215, IN
2 K.S.Rangasamy College of Technology, Tiruchengode–637 215, IN

Abstract

Group communications require a reliable multicast protocol that guarantees safe data delivery in a reasonably short time. As these applications grow in use, scalability becomes an important issue. To support these applications, several multicast transport layer protocols have been proposed and implemented. These protocols may suffer from the length of the retransmission paths between source and receivers which yields very high cost for retransmissions. An approach to reduce latency of multicast transport layer protocol is to recover the lost packet locally by requesting local replier. This paper gives the solution that local replier can be selected dynamically based on the link properties such as Bandwidth Delay Product and reliability of link. We measured recovery latency for all loss cases, namely, source, receivers, and links. Results show that the recovery latency is decreased by 17.73% to 40.59%. This percentage of improvement increases linearly, when tree size increases. Also number of duplicate replies received by the receiver is zero when loss is at the receiver and link. This approach also reduces the routing overhead significantly.

Keywords

Proxy, Bandwidth Delay Product, Recovery Latency, Exposure.

Full Text